[论文解读] Brain-Like Object Recognition with High-Performing Shallow Recurrent ANNs

Jonas Kubilius, Martin Schrimpf+12 more|Lirias (KU Leuven)|Sep 13, 2019

Domain Adaptation and Few-Shot Learning被引用 108

一句话总结

CORnet-S 是一个紧凑的四区域、带有循环的人工神经网络，在保持强劲的 ImageNet 表现的同时达到顶级 Brain-Score，强调循环性是接近大脑的关键。

ABSTRACT

Deep convolutional artificial neural networks (ANNs) are the leading class of candidate models of the mechanisms of visual processing in the primate ventral stream. While initially inspired by brain anatomy, over the past years, these ANNs have evolved from a simple eight-layer architecture in AlexNet to extremely deep and branching architectures, demonstrating increasingly better object categorization performance, yet bringing into question how brain-like they still are. In particular, typical deep models from the machine learning community are often hard to map onto the brain's anatomy due to their vast number of layers and missing biologically-important connections, such as recurrence. Here we demonstrate that better anatomical alignment to the brain and high performance on machine learning as well as neuroscience measures do not have to be in contradiction. We developed CORnet-S, a shallow ANN with four anatomically mapped areas and recurrent connectivity, guided by Brain-Score, a new large-scale composite of neural and behavioral benchmarks for quantifying the functional fidelity of models of the primate ventral visual stream. Despite being significantly shallower than most models, CORnet-S is the top model on Brain-Score and outperforms similarly compact models on ImageNet. Moreover, our extensive analyses of CORnet-S circuitry variants reveal that recurrence is the main predictive factor of both Brain-Score and ImageNet top-1 performance. Finally, we report that the temporal evolution of the CORnet-S "IT" neural population resembles the actual monkey IT population dynamics. Taken together, these results establish CORnet-S, a compact, recurrent ANN, as the current best model of the primate ventral visual stream.

研究动机与目标

• 为腹侧视觉通道建模而动机化紧凑、与大脑对齐的架构。
• 开发 CORnet-S，具有四个解剖学映射区域和区域内的循环以提高大脑相似性。
• 用 Brain-Score 与行为基准对 CORnet-S 与其他模型进行定量比较。

提出的方法

• 设计 CORnet-S，使四个区域映射到 V1、V2、V4 和 IT。
• 在区域内引入循环，且为每个区域设定具体的重复次数。
• 在 ImageNet 上使用带动量的 SGD 进行训练；批量大小 256；43 个时期。
• 使用 Brain-Score 作为综合基准，结合神经与行为数据。
• 评估神经预测性、行为预测性以及对象解决时间（OST）。
• 与包括 AlexNet、VGG、ResNet、Inception、NASNet 和 BaseNets 在内的广泛架构进行比较。

实验结果

研究问题

• RQ1 浅层、解剖对齐的循环人工神经网络在与更深层模型相比的脑样基准上的表现如何？
• RQ2 哪些架构元素（循环、瓶颈宽度、跳跃连接）对 Brain-Score 与 ImageNet 表现影响最大？
• RQ3 CORnet-S 能否捕捉到灵长类 IT 的时序动态和对象解决时间（OST）？
• RQ4 Brain-score 指标在新神经/行为数据集和迁移任务上有多好的泛化能力？

主要发现

• CORnet-S 在所测试的模型中获得最高的 Brain-Score，并在同样紧凑的模型中超越了 ImageNet 的表现。
• 循环性是对 Brain-Score 与 ImageNet Top-1 性能都具有主导预测作用的因素。
• CORnet-S 捕捉到了时序的 IT 动态，与猴子 IT 的对象解决时间相关，而不是仅限于前馈模型。
• CORnet-S 依然保持较强的 ImageNet 表现（73.1% Top-1），深度为 15，是顶尖模型中最浅的。
• Brain-Score 能在新受试者、新图像集以及 CIFAR-100 迁移测试中泛化，CORnet-S 在浅层模型中处于领先地位。